Kavitha Sundar

Q52. Explain the use of Coefficient of Variation with examples. 
 
Coefficient of variation is the ratio of standard deviation to the mean. The higher the CV, the more is the spread of the data around its mean and the team or process is very unstable or ununiformed.  In simple, it is % variation in mean, where SD is total variation in mean.  It is a measure of relative variability.
 
This is used to compare variations of two or more data sets.
 
For Eg. If I have to compare results of two groups lets say Group A & Group B. Group A has CV of  25% and Group B has CV of 18%. This says that the Group A has more variability to its mean.
 
Formula for CV = SD / mean
It can be expressed as in percentage %. Hence the formula for CV can be multiplied by 100.
 
Benefits of CV –
1.      Measure of Precision – It is used to describe the level of variations existing within the population independently from the absolute values of the individual observations. If the population is same, where you have to find out the variation, then use Standard deviation. If the population is different, use this CV to estimate the spread or variability from its relative mean.
Eg. If Male and female elephant group is compared, then use SD to find out the variation.
If you have to compare the male elephant population with male mice population, then use CV.
In simple, when the two groups differ significantly, use CV as  a measure. It is to assess the precision of the measurement technique.
 
2.      Measure of Repeatability -  CV is used to measure the repeatability within the group and not the validity / reproducibility. It is used in a way to tell you the degree of association but not agreement. Measuring repeatability with out validity is a useful analysis. When assessing the measurement error, CV value depends on both the variability between sampling units and variability between repeated readings from the same user. If we have to select the variable group of sampling units, then the repeatability CV would be higher than taking up for a homogenous group. The aim is to be maximize the repeatability within the given situation.
Eg. Used by Microbiologists and pharmacist to evaluate the intra assay and inter assay CV, in order to bring down the CV value to make it acceptable.
 
3.      Consistency of data – CV is used to understand and confirm the consistency of data. Consistency means uniformity in the values of the data set. How consistent the values are from the mean of the data set is measured. As small as the CV means the data is uniform or consistent.
Eg. If the temperature of an adult is to be compared to the same of a newborn, certain values are recorded In the real time for some time. Hence CV for adult is 10% and CV for newborn is of 2%. As for Newborn the CV is smaller, the variation in the data is very minimal. Means the data for Newborn is consistent than adult.
 
4.      Indicator for Risk Assessment – It is a better indicator for all levels of risk assessment. In any type of situation, if we were to assess the risk, this would be the right tool.
Eg. If Bank A gives a rate of interest at 20% and Bank B gives u at 10%, with a standard deviation of 10% and 5% respectively. Which bank is better to take a loan?
 As Bank B has SD of 5%, the Rate of interest is minimal for a longer run to balance his needs by the customer. Hence customer would prefer Bank B.
 
5.      Decision making: If the team has to downsize due to high cost, the decision is to eliminate some of the team members. CV Is a useful tool where it tells us in which team ,there is more of variability, which team receives higher cost , etc to make strategic decisions.
Eg. Organization has two functions – coding and billing with 40 and 65 employees in it. They earn around $450 and $350 respectively with SD as 7 and 9.
Q – A) which section has a higher salary package? Which function has highest variability?
 Answer –
a)    Salary for Coding = 40 *450 = 18000
Salary for billing = 65 * 350 = 22750
So, Salary for billing is higher.
 
    CV for Coding =( 7/450) *100 = 1.6%
CV for billing =( 9/350) *100 = 2.6%
Billing is more of variability since it has more CV.
 
The Zero disadvantage:
CV is useful only for the calculations, when the mean of sample population is not zero.
 
Lets assume, if the sample mean is equal to zero, then the denominator would become zero. Hence the CV gets nullified.
 
Yes. CV is useful if all the data points or atmost of the data points share the same value as of plus or minus sign.
 
Conclusion:
 
CV has its own use and limitations. Hence it should used to carefully in
1.      Estimating the variation 2 different populations
2.      Estimating the 2 set of categories variations.
3.      Risk assessment indicator
4.      Decision making
 
Thanks
Kavitha

Question: While continuous data is generally preferred over discrete data, please indicate circumstances where discrete is the preferred data type although continuous data is available for the same characteristic. 
 
Answer:
With given a choice on the data type, it is always useful to analyze the continuous data rather than discrete, because discrete though it has large data samples studied, the data will not be broken down into meaningful information. Continuous data can be broken down into smaller pieces and make the data informative to the decision making. With a given continuous data, we can estimate how process mean is close to or far from the target. & whether we are out of spec limits or within spec limits.
 
Example 1: (Continuous data over discrete data) Diameter of a pipe when it is produced is collected for analysis purpose. In this case, the diameter is measured in mm. Lets say, target is 10 mm and 1 mm over to it is ok. As stats are concerned, at a very high level picture, it is classified into <10mm, between 10mm & 11 mm and >11 mm. This will be projected in discrete data, as the categories/boundaries are defined and counted as defects. This has no meaning into decision making. But when the data is represented as continuous in I-MR chart, the no. of pieces which are out of spec limits are identified, root cause will be identified and arrested. It is not possible with discrete data. Hence Continuous data is always preferred.
 
Example 2: (Discrete over Continuous data)
Lets say, 20 employees working in ABC process been monitored for shift adherence. Time they login is collected against the target of shift start time & Plotted in time series chart as continuous data to find the defect %. But it will be useful in terms of RCA and not meaningful if we have to count the defect count and report out that how many were late and how many were on time.  Hence for such type of data, though the data collected from a real time scenarios and possess continuous data characteristics, it is meaningful if we present no. of late logins and shift adherence % to management as discrete data.  Here such instances like average delivery time, processing time, login time, etc falls under continuous data, for reporting purposes, it is useful to represent it as discrete data.
 
Examples where discrete data is preferred over continuous data:
 
Examples
Continuous data
discrete data
Shift adherence
Login time is noted for all employees.
For reporting purpose, continuous is converted as discrete (Late, early, on time) and presented for meaningful decisions.
Minimum balance of 1000 in bank account
Balance range is collected for all account holders.
Classified as "Maintained / Not maintained" and reported out as discrete.
Car fuel guage
how many litres remained in the car fuel tank
gauge indicates " Full, half, Empty"
Height of the child in school records
Height is noted for each and every child and compared against the growth chart wrt age of the child.
How many are underweight and overweight? Been counted from the collcted data and presented at high level.
 
 Conclusion:
Does this mean only attribute data is good enough? Of course not. Both plays a different role. For decision making, RC analysis, continuous data is more meaningful but reporting purposes at high level, discrete would be better. So answer is it depends on the underlying characteristic that we want to measure / collect and represent. If it is continuous data, then you will have the choice of reporting it out as continuous or discrete or both.
 
Thanks
Kavitha

Question: A Sponsor is someone who funds Business Excellence in a company, a strategic business unit or a function. What are the most important qualities desired in a Sponsor to ensure that Business Excellence will thrive in the organization?
 
Project Sponsor is usually a senior most person or sometimes the project manager himself or even a chair person. Resource planning is an important activity in project management. Once the resources are planned, the training takes place if the expertise skills lack. But usually, mapping happens according to the skill set of the team. Success rate of the project depends on the skills.
 
Project sponsor is also not exempted from this. A technically sound Project Sponsor should possess Problem identification and decision making skills, motivating and building cultural change in the organization as change agent, Communication both upwards & downwards, proper guidance / direction from them.
 
If the project sponsor is not willing for the project in the beginning itself, at any point in time, the project may fail. Hence it is important that the project manager gets the project sponsor involved in to project identified and aware of the milestones studied. Most importantly select a sponsor for project or else, the project suffers.
 
There are few top qualities that the sponsor should have in order to success the project. They are discussed as follows.
1.       Deep insight on the problem: Sponsor should be in a place to clearly understand the problem which has to be solved. He / She should give an answer to questions such as “What is the problem? New one / Old ? How intensive the problem is? Who created it? What is the impact? “etc. The problem may be existing problem in the organization or opportunistic approach for predictive problem.
 
2.       Deep diving into root causes: Sponsor as part of all the milestones, he should be able to collect all critical Xs Which impacts the output(Y). Identify the real root causes but not the symptoms. He should judge the analysis relating to symptoms / real root causes. For E.g. The TAT for discharging patients is around 12 hours. If the project team works on reducing the TAT from 12 hours to 4 hours as per the standard, then automating the process may be a feasible solution. But Without identifying the VA & NVA, without eliminating NVA, automation is just a wasteful activity though it adds value to the organization. It only addresses the symptoms but not the root causes. Hence process re-engineering along with automation is the best solution.
 
3.       Decision making: When a team selects a project, it is easy for them to solve multiple issues at one go with a proposed solution. If the project has to increase its scope / budget / time, a sponsor should decide considering the business requirements. Sponsor should ensure the project is focused on arresting the bleeding issues and not widely taking up all, because a team can’t do all at a time.
 
4.       Implementation of solution: Sponsor plays a major role in communicating he solutions/ changes proposed upwards and downwards. He acts as a change agent. He should clearly envision the problem, root causes, how solution helps the team, how issues will be solved if implemented, what would be the benefit, to the management.
 
5.       Basic needs, performance & Excitement needs: Sponsor needs to have this accountable to company. He should know what good enough is for the project, for the organization, for the customer. Good enough is nothing but you have taken the company’s needs, translated into requirements / CTQ and for which solution is proposed. Customer’s needs are very important while in starting and completing a project. Without delighting / satisfying the customer, a project is again not successful. Also a sponsor is cautious enough so that the organization’s resources are not wasted and should not over deliver / under deliver the solution.
 
6.       Build the team right: Sponsor should build the right team with right people, right skills, right time, etc to deliver the project deliverables. He should possess a strong relationship among the team to run the project project. He should trust the team and vice versa. He should appropriately select the team and stakeholders with the required skills. Inexperienced or new manager can’t take up the lead position, which fails the project. It doesn’t mean the new guy can’t lead. It depends on the nature of the project. Most of the time, when a project fails, it shoulders the new project manager’s/ team members. Always make them accountable for the project success. Make sure resources are rightly allocated. Shouldn’t be over staffed or under staffed. Make sure you remove if any non-performers are there in the team. This is always helpful for the success of the project.
 
7.       Accountability: Make sure as a sponsor, you make the project manager / team accountable for the success / failure of the project. Regularly meet up the team. Get the updates. Provide suggestions if required. Always keep up the timelines specified. Ensure the team is focused on the project scope and goals, timelines and deliverables. A relaxed Project sponsor will always end up with the failure of the project.
 
8.       Critical X’s Vs. Non Critical: Sponsor are the one who has to keep engaged on the issues, and the devise a strategy onto the issue. List down all the factors causing the problem, prioritize the X’s, stay focused on the critical X’s rather focusing on the non-critical ones. Sometimes proposed solution may be just enough to solve the noncritical issues. Hence sponsor should cautiously utilize the team. Being available to the bigger issue itself is a biggest help that a sponsor can do to his team. You ensure that the team is accountable for each and every task that they do.
 
9.       Mentor / Influencer: Best sponsors are not those who sits just into office and take updates once a while. Best sponsors get himself into the issue along with the team, identifies the root causes and devise a strategic plan to arrest the issue. He takes risk in trailing certain solutions. He influences people to adapt to the change prescribed. Being available is tehe best help that he can provide.
 
10.   Trial and error method: Best sponsor is the one who makes a thoughtful decisions through tough problems along with the team. He is not feared of any problems. He trails the experiments and proposes a best solution. He allows risk to become a feasible solution.
 
11.   Strong controls over the completion of the project: Sponsors always create a governance plan along with the solution to be implemented. He implements and sustains the solution in order to completely eradicate the problem from the organization. Sponsors communicate effectively the completion of the project as well being the change agent, creates a cultural change in a positive manner.
 
12.   Pull the plug / Accept the failure: Sponsor should also accept the failure not only success of the project. Uncertainty of the project to be accepted and motivate the team.
 
Summary:
Sponsor plays a vital role in business excellence model. He should be open to all challenges and active. He should be a good communicator, decision maker and strategist. He should also allow risk and help the team learn and unlearn.
Sponsors along with the team can definitely make a difference between success and failure of the project.
 
Thanks
Kavitha

Question 4 in Episode 2:
 
While continuous data is measured and attribute data is counted, there is sometimes confusion if some specific dataset should be considered continuous or attribute. Provide some examples of confusing datasets and your inference. 
 
Data – is defined as a collection of avalues / useful information that is required for any analysis to the receipient. Data is genereally used to prove / disprove hypothesis.
 
Data is of two types basis statistics. It is Quantitative or Qualitative.
Quantitative is descriptive data, which can be categorized into subgroups for analysis and qualitative is numerical which means either measurable / countable. Qualitative data is again divided into 2 types continuous and discrete data.
 
For Eg.
 
Charlie chaplin is fair, short, has small mustache, thin built and wears black colored jacket. – it is qualitative data.
 
Charlie chaplin has one hat, one walking stick and 2 legs. – it is Quantitative –discrete data.
 
Charlie chaplin aged 45 years is 57.2 kgs built and 4.8 inches tall . – it is quantitative continuous data.
 
4 types of measurement scales:
It is divided into four categories – Nominal and ordinal, interval and ratio
Ø  Nominal data: It assigns a numerical value as an attribute to any object / animal / person / any non-numerical data.
 
Ø  Ordinal data: Any data which can be ordered and ranked is called ordinal data. This can’t be measured.
Eg. 1. A horse is numbered in the race court, represents the nominal data.
2.       The numbered winning horses are ordered and ranked as “1st, 2nd  and 3rd place” in race club, which represents ordinal data. Another best examples is progress report of the student.
 
Ø  Interval: It is a numeric scale where we know order as well as the differences between values. There is no origin.
Eg. Temperature of the room is set to be normal if it is between 25 and 28 degrees C. Time is another good example of an interval scale in which the increments are known, consistent, and measurable.
 
Ø  Ratio: Ratio scales are the ultimate nirvana when it comes to measurement scales because they tell us about the order, they tell us the exact value between units, AND they also have an absolute zero–which allows for a wide range of both descriptive and inferential statistics to be applied.  At the risk of repeating myself, everything above about interval data applies to ratio scales + ratio scales have a clear definition of zero.  
Good examples of ratio variables include height and weight.
 
Qualitative data:
It is otherwise called as categorical data.
 Quantitative data:
It is divided into two contionus and discrete data.
 
Difference between Continuous and discrete data:
 
Continuous data
Discrete data
It is measureable on a scale
It is countable
The data falls within finite or infinite range
The data has only finite numbers.
Can be broken into subcategories
Can't be broken since it is a whole number.
The frequency is depicted in histogram, where skewness is shown clearly
the values take a distinct value hence it is represented in bar diagram, skewness can't be seen.
Values are allowed to group within the range
The values are individual values.
Eg. Temperature of the person, Height, Weight, Age, time, Cycle time taken to complete a task
Eg. No. of cumputeers, No. of students, no. of books, no. of certificates, no. of errors, etc
 
Confusion between Contionus and Discrete data:
 
Eg. 1:
Person
Age
Weight (Kgs)
Height(Inches)
Color
Ajay
34
51
5.1
Wheatish
Sharma
35
65.5
5.2
Fair
Roshini
23
45.5
4.8
Wheatish
Gaithri
53
72.5
4.8
Dark
Linda
43
46.5
5.1
Fair
Tanya
36
43
5.3
Wheatish
Balu
27
56
5.6
Fair
Vignesh
32
77
6.1
Dark
Aarav
43
76
5.9
Wheatish
Rithesh
45
64
5.3
Dark
 
Qualitative data / categorical data:
Categorize 10 people in the group into wheatish, dark, fair basis the color. This represents categorical data.
 
Continuous data:
Age , Height and weight of the people displayed above in the table depicts a good example of continuous data, where these numbers falls within the infinite ranges.
 
Discrete data:
No . of Wheatish – 4
No. of fair – 3
No. of dark – 3
Total no. of people – 10
 
Conclusion of Eg. 1: Though age is continuous numerical variables. Although the recorded ages have been truncated to whole numbers, the concept of age is continuous.) Number of aged people is a discrete numerical variable (a count).
 
Age can be rounded down to a whole number, if so it represents the discrete data. Though it falls under discrete(when all data is shown as whole integers), it is actually a continuous data because it has ranges. Age is not a constant factor, though the DOB is constant.
 
Basis the context / concept of the requirement – lets say to fill a form, the exact age is required. In such case, though age is discrete, it is continuous.
 
“12 years, 153 days” really means a continuous age that is between 12Y152.5D and 12Y153.5D.”
 
Eg. 2 : Income is another example of continuous data.
 
Eg. 3: “
 In practice, percentage data are often treated as continuous because thepercentage can take on any value along the continuum from zero to 100%. In addition, dividing a percentage point into two or more parts still makes sense.Discrete data are easy to collect and interpret.
 
% is always to be considered as continuous but it depends on the concept.
 
If I have to track the error percentage, the right metric is as below..
 
Error % =          No of errors (Discrete)       
                  Total charts audited.(Discrete)
 
Hence Error % is discrete.
 
Another example:
 
If I have to track the availability of the machine, the formula is as follows…
 
Availability % = Total hours available (Continuous) / Expected hours of production for 8 hours(Continuous)
 
Hence Availability % is continuous, since time is continuous.
 
 
Conclusion:
 
It depends….  In certain situations, discrete data may take on characteristics of continuous data.  But, if counts are large, distribution of values are relatively wide, and the the values are distributed across the values, you can “pretend” it is continuous and use the appropriate tools.
 
Thanks
Kavitha